Maybe in some jurisdictions, but I’m pretty sure many places exclude copies necessary for the operation of the software to not be violations of copyright law.

I mean, extend the same reasoning to web pages. Your web browser downloads a web page (let’s pretend this does not count as a copy). It keeps the HTML in an in-memory cache. The browser then sends the HTML to its internal renderer, which renders the page. Boom, a copy of the HTML (or at least a derived work of it) now exists in the renderer. Have you now violated the copyright of the page author?